W3C home > Mailing lists > Public > www-svg@w3.org > January 2010

Re: SVG Presentational Attributes, attributeType="CSS" vs "XML": same or different?

From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
Date: Fri, 22 Jan 2010 11:55:24 +0100
To: www-svg@w3.org
Message-Id: <201001221155.24902.Dr.O.Hoffmann@gmx.de>
The relevance of attributeType has a long discussion history ;o)

In doubt for details SVG animation refers to SMIL and this
explains in the sandwich model, what applies.
Historically this was maybe intended to be simpler, however
when I was a member of the SVG group, I tried to clarify this
and with some discussion of the issue with the SYMM group
they confirmed again, that it is of some relevance in the sandwich 
model and the interaction of animation of XML presentation 
attributes, CSS stylesheet properties and animation of CSS 
properties results in a complex sandwich of priorities.
Maybe this could have been solved much simpler and more
friendly for authors, when SMIL animation appeared, for 
example with some pseudo namespace for CSS, but this
did not happen and I think, now it is a little bit late to change
this (maybe unfortunately).

Apart from a few errors the results with Opera for my group of
related tests are indeed the best, what means, that some versions 
of Opera get more tests right than all other tested viewers.

I cannot see a problem with this:
 > Animation of presentation attributes is equivalent to animating the
 > corresponding property. Thus, the same effect occurs from animating
 > the presentation attribute with attributeType="XML" as occurs with
 > animating the corresponding property with attributeType="CSS".

Of course, there is itself no different effect, if it is interpreted 
correctly, what 'effect' means). Obviously the visible result for an
animated attribute or property depends on many other things, 
on the underlying  value and priorities of different animations,
stylesheets etc. The animation function (See SMIL) is the same 
in the sense, that the computations are the same, not necessarily 
the final presentation result for a real graphical object at some 
specific time. Finally, directly before presentation on has to join
together all these applicable presentation attributes, properties,
stylesheets and animation functions in a proper order with the
right priorities to be able to present the graphical object.
In my tests and discussions it turned out, that this is often 
implemented wrong (for example for additive, cumulative
behaviour and especially for to-animations) already without the 
additional problem of the attributeType, therefore it is no surprise, 
that one finds even more errors in current implementations, if 
the situation is complex enough, that attributeType becomes
relevant ;o)


If one manages to isolate a simple enough sample,
the visible result can be the same. Of course, if there are other
dependencies and priorities, this can be quite different, this 
happens with and without attributeType. And what applies is
defined in the sandwich model in SMIL (note that there are
minor clarifications in SMIL3 due to some discussion of this
nasty problem).

Olaf
Received on Friday, 22 January 2010 11:05:28 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:44 GMT